Prostate cancer (PCa) is a leading cause of cancer morbidity and mortality in men, and the androgen receptor (AR) is the primary therapeutic target. In early PCa, anti-androgen therapy (AAT) is almost universally effective. This consists of one or more combinations of GnRH agonists (to suppress pituitary signaling), aromatase inhibitors (to decrease androgen production), and competitive AR antagonists (to block AR directly) such as hydroxy-flutamide (OH—F) or bicalutamide (BiC). This strategy usually works for several years, but over time tumor cells evolve mechanisms for continued growth under these conditions of androgen depletion. Most recurrent, or hormone-refractory prostate cancer (HRPC) is nonetheless dependent on AR-mediated signaling. This can include upregulation of AR protein expression levels, acquisition of mutations within AR that increase its activity in response to alternative hormones (including antagonists), or upregulation of co-activator proteins that augment AR activity. Thus, it is likely that new approaches to block AR activity could significantly extend or increase the effectiveness of AAT. This could consist of better competitive antagonists, and considerable efforts from pharmaceutical companies are already being brought to bear on this approach. For example, pyrvinium has shown efficacy against tumor cell proliferation in vitro and in mouse models (see WO2006/078754 and Cancer Cell 2006, 10, 321). This implies that novel anti-androgens might have considerable utility in the treatment of both primary and recurrent PCa. Such anti-androgens might not be competitive antagonists that directly bind AR, and could conceivably function via inhibition of downstream events in AR signaling. Accordingly, there is a need for novel, potent anti-androgens. Surprisingly, this invention meets this, and other, needs.